Geo opt with fixed atoms relative positions

[HB H]

Dear users,

I want to optimize the structure with two atoms distance fixed. However the distance between tow atoms (index 44 and 575) still changes during the optimization. 

I know by using "&FIXED_ATOMS" can freeze atoms, but I just want to fix the relative distance between atoms, not overall atomic position.  The input is shown below and it only works for MD? Thx!

&GLOBAL

  ...

  RUN_TYPE GEO_OPT

 ...

&END GLOBAL  

&FORCE_EVAL

.....

    &SUBSYS

.....

    &COLVAR

       &DISTANCE

         ATOMS 44 575

         AXIS XYZ

       &END DISTANCE

    &END COLVAR

        

  &END SUBSYS

&END FORCE_EVAL

&MOTION

  &GEO_OPT

    TYPE MINIMIZATION

    MAX_DR    1.0E-03

    MAX_FORCE 1.0E-03

    RMS_DR    1.0E-03

    RMS_FORCE 1.0E-03

    MAX_ITER 2000

    OPTIMIZER BFGS

  &END GEO_OPT

&CONSTRAINT  

    &COLLECTIVE

      COLVAR 1

      TARGET  xxxx

      INTERMOLECULAR T

    &END COLLECTIVE

  &END CONSTRAINT

&END MOTION

[jjh...@gmail.com]

Hello,

This method will only work for MD calculations. You can try using a restraint (  MOTION / CONSTRAINT / COLLECTIVE / RESTRAINT ) with a high K value.

Best,

Julian

[HB H]

Thx Julian,

I have changed the  &CONSTRAINT section as suggested, and it works for me.

&CONSTRAINT  
    &COLLECTIVE
      COLVAR 1
&RESTRAINT
K=5.0
&END RESTRAINT

      INTERMOLECULAR T
    &END COLLECTIVE
  &END CONSTRAINT

Cheers,

Haibei

jjh...@gmail.com <jjh...@gmail.com> 于2020年9月15日周二 下午6:12写道：

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/a215ee25-1321-4f82-88fb-597324e8e03cn%40googlegroups.com.